Edinburgh: FPGA core for high performance sequence alignment

Researchers at Edinburgh University have developed a Field Programmable Gate Array (FPGA) core that they say provides a high performance and a versatile and accurate way of carrying out comparative analysis of large sequence datasets.

Sequence alignment involves the arrangement of data sequences in order to identify and provide important information on key regions of similarity. Such pattern-matching is important in the analysis of biological data, natural language processing and stream data processing within network systems.

This FPGA core is faster than equivalent ‘software only’ approaches and uses only a fraction of the power. It would therefore be an effective tool for operations normally requiring a cluster of computers.

In comparison to current FPGA-based parallel technologies for sequence alignment, the core is highly parameterised, allowing user definition of processes. It is claimed to run 200 times faster than standard software systems.

A hardware version of an FPGA-based web server for sequence alignment of biological data has been successfully demonstrated.

Edinburgh University is seeking commercial organisations interested in licensing the FPGA core on a non-exclusive basis.